Ask HN: Does doing what your users say they want ultimately end in betrayal? | Hacker News Reader

Ask HN: Does doing what your users say they want ultimately end in betrayal?

4 pointsdiminium4 comments

You know the story. You listen to your users and do exactly what they say they want. You implement the features they want and you always keep their input in the process. From polls to surveys to comments - everything you make goes by the user. Sounds good in the surface, right?

That is until some company who completely ignores their user and does what they think is correct comes along (say Apple). Now the users you've been taking care of since the beginning have left your products in droves.

What happened? Why did they leave? You did everything they said but they still left. What's going on here?

View on HN

My perspective is you should listen to what your users say, but treat that as just one piece of input into what you do.

Ie, rather than this:

what we do = what users tell us

It should be:

what we do = f(what users tell us, x)

Where f is your own thinking, and x is other sources of useful data.

My feeling is that "vision" and "data" aren't in conflict -- you should attempt to integrate data into an over-riding vision, but you should be prepared to ditch or tweak the vision if it becomes too detached from reality.

I could write a lot more about this -- I think the problem goes a bit deeper than "faster horse" dynamics that some people have mentioned. The whole lean startup, listen to your customers thing is supposed to be about applying the scientific method to founding companies.

Trying out random things and recording the results might be "experimenting" in the normal sense, but it's not the scientific meaning of experimenting. You should concoct a theory and work out what you need to look for before you start looking. The Lean Startup book itself goes into a lot of depth about which metrics to track, how different marketing models require different metrics, etc (everyone thinks they know "lean concepts" because they read a few blog posts but the book goes into way more depth).

Can't say I can name a company that has happened to. Do you have any examples in mind?

Here are a couple of ideas.

If you ask users how to make X better they will suggest incremental changes to X. When you phrase the question "How do we improve the current solution to a problem" you rarely get a different solution. You quite often get an incremental improvement to the solution. Notice how you have quit focusing on the problem, and started focusing on your solution? It rather binds what you are willing to consider in the future.

The second is doing whatever your customers want is even worse than design by committee, it is design by a committee that doesn't interact or have the same general goals. A Feature set should provide a clear an concise vision. Committees rarely have that, and random people suggesting things certainly don't.

Sometimes users are unaware of what they want and it takes another company (Apple) to come along and show them what they really want.

Steve Blank had a fantastic post about this you might want to take the time to read.

http://steveblank.com/2012/02/27/killing-your-startup-by-lis...

Reply on news.ycombinator.com